DNA polymerase that replicate chromosomal DNA are highly processive: they can replicate thousands of nucleotide with dissociation. The goal of the proposed research is to obtain structures of the subunits of these polymerase and to deduce the principles underlying their high speed and fidelity. The immediate objective is to determine the structures of two proteins for which suitable crystals are in hand: the delta, subunit of E. coli DNA polymerase III (Pol III) and the DNA clamp of eukaryotic DNA polymerase delta, PCNA (proliferating cell nuclear antigen) from the yeast Saccharomyces cerevisiae. DNA replication and repair is at the heart of error-free cell division, and a complete understanding of these processivity factors has important implications for understanding the origins of mutations and resulting disease. We have previously determined the structure of the beta subunit of E. coli DNA polymerase III, which clamps the polymerase by forming a closed ring which can encircle DNA. The ring is assembled on DNA by the gamma complex (5 proteins: gamma,delta, delta, chi, psi) and iota subunit. In collaboration we have crystallized the delta subunit, and propose to determine its structure by multiple isomorphous replacement. Due to sequence homology, the structures of gamma and iota will be partially deduced from the resulting structure of delta. Since iota and delta alone suffice to form a functional (but less active) clamping machinery, the structure of delta will provide insights into the mechanism. The crystals diffract to 2.7 A on a rotating anode generator (orthorhombic C-centered, a=99.8, b=105.0, c=76.9A). The protein that corresponds to E. coli beta-subunit in eukaryotic systems is PCNA which is also required for DNA excision repair, and interacts with cell cycle proteins. In collaboration yeast PCNA crystals that diffract to 2.4A on a rotating anode X-ray generator at cryogenic temperatures (cubic P213, a=121.1 A) have been obtained. The best diffraction is observed for a non-isomorphous mercury derivative, with a strong anomalous signal, and this will be used to obtain phases by multi-wavelength anomalous diffraction. Other work: Preliminary crystals have been obtained of T4 phage gene 45 protein (diffraction demonstrated to 3 A), and of human PCNA (not tested), autoantibodies against which are implicated in systemic lupus erythematosus. Conditions for their growth will be optimized. All the 10 protein subunits of E. coli Pol III are available to us in pure form for crystallization, and we propose to see if they can be crystallized in various combinations. We have already established that the catalytic alpha subunit (130 KDa) does not readily crystallize, and we propose to study its proteolytic fragmentation patterns with the intention of discovering sub-structures that are more suitable.